Predicting the Risk of Cardiovascular Diseases using Machine Learning Techniques

Abstract

These days, health-related diseases are increasing day by day due to lifestyle and genetics. Especially these days, heart disease is so common that people's lives are at risk. Blood pressure, cholesterol and pulse rate vary from person to person. However, according to proven clinical results, normal blood pressure is 90/120 and cholesterol is 129-100 mg/dL, Pulse 72, fasting blood glucose 100 mg/dL, heart rate 100-60 bpm, normal ECG, main vessel width 25 mm (1 inch) in the aorta only 8 μm in the capillaries. This article looks at the different classification techniques used to predict each person's risk level based on age and gender. Blood pressure, cholesterol, heart rate. A "disease prediction" system based on predictive modeling predicts a user's disease based on the symptoms the user enters into the system. The system analyzes the symptoms that the user provides as inputs and provides disease probabilities as outputs. Disease prediction is done by applying techniques like KNN, Decision tree classifiers, random forest algorithms, and more. This technique calculates the probability of a disease. Therefore, we obtain an average prediction accuracy probability of 86.48%.

Full text

International Journal of Intelligent Systems and Applications in Engineering
IJISAE, 2023, 11(2s), 165–173 | 165
International Journal of
INTELLIGENT SYSTEMS AND APPLICATIONS IN
ENGINEERING
ISSN:2147-6799 www.ijisae.org Original Research Paper
Predicting the Risk of Cardiovascular Diseases using Machine Learning
Techniques
Puneet Garg1, Neetu Sharma2, Sonal3, Bharati Shukla4
Submitted: 22/10/2022 Accepted: 24/01/2023
Abstract: These days, health-related diseases are increasing day by day due to lifestyle and genetics. Especially these days, heart
disease is so common that people's lives are at risk. Blood pressure, cholesterol and pulse rate vary from person to person.
However, according to proven clinical results, normal blood pressure is 90/120 and cholesterol is 129-100 mg/dL, Pulse 72, fasting blood
glucose 100 mg/dL, heart rate 100-60 bpm, normal ECG, main vessel width 25 mm (1 inch) in the aorta only 8 μm in the capillaries.
This article looks at the different classification techniques used to predict each person's risk level based on age and gender. Blood
pressure, cholesterol, heart rate. A "disease prediction" system based on predictive modeling predicts a user's disease based
on the symptoms the user enters into the system. The system analyzes the symptoms that the
user provides as inputs and provides disease probabilities as outputs. Disease prediction is done by applying techniques like KNN,
Decision tree classifiers, random forest algorithms, and more. This technique calculates the probability of a disease. Therefore, we obtain
an average prediction accuracy probability of 86.48%.
Keywords: Coronary ailment, Heart disease, ECG, Data mining, Respiratory disappointments
1. Introduction
In regular daily existence there are various parts that
impact the core of individual viable. Various issues are
going on at a disturbing velocity and new heart
contaminations are immediately perceived. Heart is the
fundamental blood pumping organ; it is responsible for
circulating blood in the entire body and its smooth
functioning is very important for a person’s survival. A
person’s health especially heart health is widely
dependent on a person’s lifestyle, his behavior and
profession. Heart disease is dependent on a genetic factor
of a person, that is his genetic makeup which he has
acquired from his ancestors. There are various Genetic
factors which are passed from generations through which
various types of heart diseases are passed on. According
to the World Health organization’s study all over the
world more than 13.99 million deaths are occurring and
the cause of death in these cases is identified as
cardiovascular diseases, these diseases affect the arteries
of a heart and thus affect the circulating power of a heart.
Younger generations in their twenties are also being
affected due to cardiovascular diseases. Lifestyle of a
person affects the possibility of being affected by heart
diseases, obesity is such a cause, eating junk food makes
a person obese and increases possibility of heart disease.
Poor diet, not consuming a balanced diet also leads to
heart disease [20,21]. High blood pressure also leads to
heart diseases. Smoking is injurious to health in every
aspect and it leads to heart disease as well [27,28,29,30].
Below figure 1. Describe how smoking is affected on
human heart and which type of problem human have
been faced.
1Assistant Professor, ABES Engineering College, Ghaziabad,
U.P., India
2 Professor, Galgotias University, Greater Noida, U.P., India
3Associate Professor, Department of CSE, BPSMV, Khanpur
kalan, Distt. Sonipat, Haryana, India
4Assistant Professor, ABES Engineering College, Ghaziabad,
U.P., India
puneetgarg.er@gmail.com1,neetush75@gmail.com2,
sonalkharb@gmail.com3, bharati.shukla@abes.ac.in4

International Journal of Intelligent Systems and Applications in Engineering
IJISAE, 2023, 11(2s), 165–173 | 166
Fig. 1. Smoking effects on heart
After being affected by a heart disease, there should be a
proper mechanism for diagnosing the disease especially
the heart disease, it’s the most complicated task in the
health care system. Various factors are taken into
analysis while predicting the nature of a diseases in our
case heart diseases, above mentioned lifestyle factors and
others habits are taken into consideration. While
performing checkup of a patient, doctor in charge tries to
figure out all the factors. The symptoms of coronary
ailment exceptionally depend on which of the misery felt
by a man. A couple of reactions are not commonly
recognized by the standard residents. Nevertheless,
ordinary signs join chest torment, brevity of breath, and
heart palpitations. The chest torment fundamental to
various sorts of coronary sickness is known as coronary
thrombosis and generally takes place due to shortage of
oxygen in the body. Severe chest pain may be actuated
by disagreeable exercise and consistently suffers under
5-15 minutes [38,39]. Respiratory disappointments can
in like manner happen Coronary ailment Figure using
simulated intelligence Figuring’s in view of different
sorts of coronary sickness. The signs of a cardiovascular
disappointment take after chest discomfort. The signs of
a coronary disappointment can a portion of the time take
after indigestion. Indigestion and gastrointestinal
discomfort is a possible symptom. Diverse indicator of a
respiratory disappointment fuse torment that develops in
a person, it appears as a pain in firearms, cervix, midriff ,
belly, or mandible, deliriousness what's increasingly,
woozy sensations, copious sweating, infection and
heaving.
Fig. 2. Cardiovascular System
Figure 2. represent the parts of Cardiovascular system.
The Cardiovascular breakdown [10,11] is moreover an
aftereffect of coronary ailment, and windedness can
happen when the heart ends up being too weak to even
consider evening consider circling blood. Some heart
conditions occur with no reactions using any and all
means, especially in increasingly in persons with a
history of polygenic disorder. Heart diseases present
from birth have some general characteristics such as high
perspiration levels, lethargy, rapid heartbeat and
abnormal breathing[46,47]. In these circumstances, the
investigation transforms into astounding undertaking
with an unimaginable acquaintance with inclination. A
respiratory[44] disappointment or the possibility of the
cardiovascular problem at whatever point separated
early, can empower the patients to avoid any and all risks
and take managerial measures. Starting late, the human
administrations industry has been making tremendous

International Journal of Intelligent Systems and Applications in Engineering
IJISAE, 2023, 11(2s), 165–173 | 167
proportions of record of information of people affected
or contamination discovering reports are all things
considered especially taken for the desire ambushes far
and wide. Right when the data about coronary sickness is
enormous, the man-made intelligence systems can be
executed for the assessment. In this investigation work,
the managed computer knowledge is taken into
consideration for getting accurate results for analysis and
predicting things in advance [41,42][45]. The knowledge
is not visible to everyone and can be used only by
utilizing proper data mining tools. Thus, there arose need
for classifying data mining into various types through
which successful results may be obtained which can be
utilized in medical field. Thus, gave birth to medical data
mining which combined various techniques and gave
computer-based schooling to the provided input or we
can say datasets thus leading to possible exploring of
various hidden systems in the medical data mining which
predicted the patients’ symptoms. Thus, by making use
of medical data mining we can generate accurate results
for focusing on the cause of illness and predicting actual
disease. Medical data mining is utilizing various
flowcharts and algorithms to provide justice to the
investigation. We can make use of various algorithms to
determine the type of heart disease social protection is a
field of the most required assistance and a monetarily
2ndlargest industry in 21stcentury. While we talk about
the sensibility and quality insistence in human
administrations industry, a couple of quantifiable
examinations are kept making prosperity game plans
logically precise and impeccable in this rhythmic
movement time of extending clinical issues and
consistent diseases. Types of progress on data driven
insightful headways is contamination finding and
acknowledgment, treatment and research are striking.
Clinical picture assessment, reaction based
contamination desire is the place the most searched for
after brains are working [22,23,24]. In this paper we
intend to present our proposed model on the gauge on
finish of cardio vascular infirmity with ECG assessment
and sign based distinguishing proof. The model hopes to
be investigated and advance in further to get solid and
from beginning to end reliable research instrument. We
will look at about the customary procedures and
computations completed on CVD conjecture, dynamic
degrees of progress, draw assessment of execution
among the present structures and propose a redesigned
multi-module system performing better similar to
accuracy and plausibility. Utilization, getting ready and
testing of the modules have been done on datasets
procured from UCI and Physio net data stores. Data
position have been adjusted if there ought to emerge an
event of the ECG[26] report data for development of
movement by the convolutional neural framework used
in our investigation and in the peril figure module, we
have picked properties for getting ready and executing
the multi-layered neural framework made by
us[36,37,38,][4].
The structure of remining paper is as follow: Section II
describe the detail literature part of cardiovascular
system. Section III discussed the problem statement of
research paper and the datasets which are required to
implement problem statement are discussed in Section
IV. Result of research paper are discussed in Section V
and conclusion of our paper are discussed in Section VI.
2. Literature Review
As per Shantakumari, et al. [1] have achieved an
assessment work in which the adroit and sensible
cardiovascular disappointment figure structure is made
using Multi-Layer Perceptron with Back-Duplication. In
like way, the intermittent events of the coronary
suffering are assembled with the MAFI A estimation
subject to the values evacuated.
Yanweeii, et.al [2] ,Yadav. et.al [25] have accumulated a
social event framework reliant on the reason behind
parametric specifications by researching HRV (Heartbeat
Changeability) from electrocardiogram and the data is
earlier managed with coronary ailment surmise model is
made that coordinates the coronary issue of a sick
person. A couple of data mining methods are used in the
assurance of coronary sickness, for instance, Guiltless
Bayes, Decision Tree, neural framework, partition
thickness, sacking computation, and reinforce vector
machine exhibiting different degrees of exactnesses.
Blameless Bayes is amongst the compelling portrayal
frameworks used in the assurance of coronary sickness
patients.
Die doownn et al. [3] discussed about another part of
decision method count which is the cream methodology
which merged Bayes speculation and surveyed estimated
accuracy 85.5559%. Regardless of the way that these are
commonly used artificial intelligence computations, the
coronary disease gauge is an essential task including
most vital possible exactness. Accordingly, the three
figurings are checked at different stages and sort to
appraisal philosophies.
Karayilan et al.[5] and Mardiyono et al.[6] discussed a
brief about clinical experts in developing an unrivaled
appreciation and help them with recognizing a response
for perceive the best technique for foreseeing the heart
diseases.
Guru et al.[7] and Singh et al.[8] developed a key test
confronting social protection affiliation (crisis facilities,
clinical centers) is the workplace of significant worth
organizations at reasonable expenses. Quality solaces

International Journal of Intelligent Systems and Applications in Engineering
IJISAE, 2023, 11(2s), 165–173 | 168
propose detecting sickness exactly and preventing
prescriptions that are practical. Inadequate clinical
choices can impel disastrous results, that are as such
unacceptable. Crisis centers ought to compel the
expenses on performing these clinical tests. This can be
achieved by utilizing fitting PC type of data and in
addition choice really solid frameworks.
Asha et al. [9] studied on the function of the heart isn’t
genuine, it will have a direct imapct on working of other
body parts, for example, cerebrum, nephrons, and so on.
Coronary sickness is a tribulation which ramifications
for the development of the heart.
Nasira et al. [10] and other researchers [11][35]
discussed a wide accessibility of champion extent of
information and a need to change over this open epic
extent of information to steady data requires the
utilization of information mining philosophies.
Information Mining and KDD (learning disclosure in the
database) have wound up being prominent as of late.
Daniel et al. [12] discussed on the omnipresence of
information mining and KDD approach (information
divulgence in database). It shouldn't be a marvel since
the extent of the information constructs that are
accessible are incredibly wide to be dismembered truly
and even the techniques for redid information assessment
considering set up bits of information and machine
modifying a great part of the time bargain issues while
arranging immense, amazing information grows
including complex things.
Kumari et al. [13] discussed on information Mining
approach and it is the feature of Data Exposure Database
(KDD). There are sure periods of information mining
that you ought to get settled with, and these are
examination, plan recognizing evidence, and course of
action. Information mining is an iterative technique that
by and large consolidates the going with stage.
Mehta et al. [14] and X et al.[15] discussed the
associations of disarranges and the certified purposes
behind the messes and the effects of reactions that are
abruptly found in patients can be surveyed by the
customers by methods for the manufactured
programming with no issue.
Jankowski et al.[16] and other researchers [17,18] does
accurate examination has recognized the disperses of the
heart and veins, and joins coronary sickness (respiratory
disappointments), cerebrovascular contamination
(stroke), raised circulatory strain (hypertension),
periphery course disease, rheumatic coronary ailment,
inherent coronary ailment and cardiovascular
breakdown. The critical explanations behind
cardiovascular disease are tobacco use, physical
inertness, an awful eating daily schedule and frightful
using ethanol. The three huge purposes behind heart
sickness are distress, brain damage and coronary
disappointment.
Singh et al. [19] implemented a K-suggests approach and
the knowledge extraction methodologies like fake neural
methodology utilized in amazing coronary
disappointment desire figure of heart contaminations was
pre-arranged and assembled by strategies for K-suggests
packing count. By then neural framework is set up with
the picked basic models.
3. Problem Definition
Huge increase in number of patients with heart diseases
has been pointed out and this is as a result of our eating
food and then moving back to workplace without doing
any physical exercise. Our inactive lifestyle is largely
responsible for this increase. Technology has played a
vital role for this change in our behaviour with the
availability of cell phones people prefer to utilize their
leisure time watching movies, playing games or various
our inactive recreational activities. Our generation is too
lethargic to take a walk, or do some kind of
meditation[31,32,33,34] . This inactive lifestyle has led
to an unfortunate increase in number of heart related
sickness. India being a developingcountry people mostly
perform 9-5 kind of jobs which is an inactive lifestyle
and thus it has the same effect on increase in
Cardiovascular illness. People nowadays use their own
transport even for shorter distances, For buying vegies
people prefer to use their car or other transport available
rather than waking some distances. The death rate per
200,000 people from Heart [40][43][45].
In this research paper, 3 algorithms is used and have
formulated a model to get accurate results in predicting
the heart diseases present in an individual. The dataset is
used from UCI repository.
3.1.
Dataset, Dataset Structure & Description
To make our research realistic we have taken datasets
from Cardiology Institute from Hungary university. We
have used dataset from UCI repository.
3.1.1
Different Stages of datasets involved for
reached our proposed approach
The Figure 3. Represent overall stages required to
reached our problem objective i.e. to calculate accuracy
of different classification of machine learning techniques
used for predicting the risk level of each person based on
age, gender, Blood pressure, cholesterol, pulse rate.

International Journal of Intelligent Systems and Applications in Engineering
IJISAE, 2023, 11(2s), 165–173 | 169
Fig. 3. Stages of datasets
3.1.2. Initialization of datasets to achieved problem
statement are as follow
Stage 1: Import Libraries
First, declare java libraries for comparing the
classification algorithm. Figure 4 represent the java
libraries which are used to defined datasets. For
comparison and analysis of classification algorithm the
proposed approach used Cardiology datasets.
Fig. 4: Import libraries
Stage 2: LOAD The Data
After initialize or declare the library, now second step is
importing the data file. Below figure 5. represent the
syntax which are used for initialized the data file. The
datasets are taken from Cardiology Institute for analysis
the classification algorithms.
Fig. 5. Load the datasets
Stage 3: Check the type of the dataset
After initialize the data file with the help of syntax
check the type of dataset. By using dataset accuracy is
calculate of classification algorithms. Figure 6. represent
the syntax which is used to declare type of datasets.
Fig. 6. Type the datasets
Stage 4: Check the Shape of the data
After initialize the data file with the help of syntax check
the shape or size of dataset. By using dataset Cardiology
Institute comparison of machine learning classification
algorithm has to be performed. Figure 7. represent the
syntax which is used to declare shape of datasets.
Fig. 7. Shape of the datasets
Stage 5: Dataset description
After declare shape and type of dataset finally in this
stage, datasets are found in detail form i.e., age, sex,
target, slope, cp, fbs, etc. Then datasets is applied to
classification algorithm for analysis the accuracy for
disease prediction purpose. Figure 8. represent the syntax
which is used to declare datasets in detail.

International Journal of Intelligent Systems and Applications in Engineering
IJISAE, 2023, 11(2s), 165–173 | 170
Figure 8. Dataset in detailed form
Stage 7: Check for missing Data
After getting data in describe form check wheatear any
kind of data is missed at the time of declaration. Below
Figure 9. Represent syntax which is used to check data is
missing or not.
Figure 9. data is missing or not.
Stage 8: Check the correlation with target data
After getting data in describe form check data is
correlated with our target or not . Below Figure 10.
Represent syntax which is used to check data is
correlated.
4. Results and Discussions
Fig. 10. data is corelated with target
purposed of applying algorithm to predict the disease in
As per discussion in Section III, to meet our objective
the Cardiology Institute datasets is used which taken
Hungary University. After declare or initializing
datasets in describe form apply dataset on some machine
learning classification algorithm they are KNN, Decision
Tree Classifier and Random Forest etc. The main
human body i.e., blood pressure, cholesterol, pulse rate,
etc. Figure 8. Shows the comparison graph of
classification algorithm on the basis of accuracy
parameter. After observation of Figure 8. it assumed
average prediction accuracy probability 86.48% is
obtained.

International Journal of Intelligent Systems and Applications in Engineering
IJISAE, 2023, 11(2s), 165–173 | 171
5. Conclusion
Fig. 8 Accuracy Scores of machine learning algorithm
retraining.
The general target of our task is to foresee precisely with
less number of tests and characteristics the nearness of
coronary illness. In this task, fourteen properties are
viewed as which structure the essential reason for tests
and give exact outcomes pretty much. A lot more info
properties can be taken however we will likely anticipate
with less number of credits and quicker effectiveness to
foresee the danger of having coronary illness at a
specific age range. Five information mining order
methods were applied to be specific K-Closest Neighbor,
Innocent Bayes, Choice Tree, Irregular Woodland and
Calculated Relapse. It is demonstrated that Arbitrary
Timberland has preferable exactness over different
systems.
This is the best model to anticipate patients with
coronary illness. This undertaking could answer complex
questions, each with its own quality effortlessly of model
understanding, access to nitty gritty data and precision.
This undertaking can be additionally improved and
extended. For instance, it can fuse other clinical traits
other than the 14 qualities we utilized. It can likewise
join other information mining systems, e.g., Time
Arrangement, Bunching and Affiliation Rules. Constant
information can likewise be utilized rather than simply
all out information. Another region is to utilize Content
Mining to mine the immense measure of unstructured
information accessible. This undertaking is introduced
utilizing information mining procedures. From calculated
relapse, KNN, Innocent Bayes, Choice Tree, Irregular
timberland are utilized to build up the framework.
Arbitrary Timberland demonstrates the better outcomes
and helps the area specialists and even the individual
identified with the clinical field to get ready for a
superior and early analysis for the patient. This
framework performs reasonably well even without
References
[1]
S. F. Dessai, “Intelligent Heart Disease Prediction
System Using Probabilistic Neural Network,” no. 5,
pp. 38–44, 2013.
[2]
E. Y. and C. Kilikcier, “Determination of Patient
State from Cardiotocogram using LSSVM with
Particle Swarm Optimization and Binary Decision
Tree,” Master Thesis, Dep. Electr. Electron. Eng.,
no. Uludag, 20.
[3]
N. S. and D. Singh, “Performance Evaluation of K-
Means and Hierarchal Clustering in Terms of
Accuracy and Running Time,” Dep. Comput. Sci.
Eng. Barkatullah Univ. Inst. Technol., no. Ph.D
Dissertation, 2
[4]
I. S. F. Dessai, “Intelligent Heart Disease Prediction
System Using Probabilistic Neural Network,” no. 5,
pp. 38–44, 2013.
[5]
T. Karayilan and Ö. Kiliç, “Prediction of Heart
disease using neural network,” in 2nd International
Conference on Computer Science and Engineering,
UBMK 2017, 2017, pp. 719–723.
[6]
M. Mardiyono, R. Suryanita, and A. Adnan,
“Intelligent Monitoring System on Prediction of
Building Damage Index using Neural-Network,”
TELKOMNIKA (Telecommunication Comput.
Electron. Control., vol. 10, no. 1, p. 155, 2015
[7]
N. Guru, A. Dahiya, and N. Rajpal, “Decision
support system for heart disease diagnosis using
Neural Network,” 2007.
[8]
J. S. Singh Navdeep, “No Title,” Intell. Hear. Dis.
Predict. Syst. using CANFIS Genet. Algorithm, Int.
J. Biol. Med. Sci., no. International Journal of
Advance Research, Ideas and Innovations in
Technology© 2018, www.IJARIIT.comAll Rights,
p. Reserved Page | 987, 2008.
100.00%
90.00%
84.25%
86.89%
81.96%
84.25%
80.00%
70.00%
60.00%
50.00%
40.00%
30.00%
20.00%
10.00%
0.00%
68.85%
KNN
Decision tree Logistic Naïve bayes Random
regression Forest
ACCURACY

International Journal of Intelligent Systems and Applications in Engineering
IJISAE, 2023, 11(2s), 165–173 | 172
[9]
R. G. S. Rajkumar Asha, “No Title,” Diagnosis
Hear. Dis. Using Datamining Algorithm, vol. Issue
10 V, no. Global Journal of Computer Science and
Technology, p. Page 38, 20
[10]
S. Radhimeenakshi and G. M. Nasira, “Prediction
of Heart Disease using Neural Network with Back
Propagation,” Int. J. Data Min. Tech. Appl., vol. 4,
no. 1, pp. 19– 22, 2016.
[11]
“Prediction of Heart Disease using Artificial
Neural Network,” VFAST Trans. Softw. Eng., pp.
102–112, 2019.
[12]
Halaudi Daniel M., “Prediction of heart disease
using classification algorithms.,” WCSECS, pp.
22–24, 2014
[13]
G. S. Kumari Milan, “No Title,” Comp. Study Data
Min. Classif. Methods Cardiovasc. Dis. Predict.,
vol. Vol. 2, no. IJCST Vol. 2, Iss ue2, June 2011,
2011
[14]
Z. M. A. Mehta Rupa G. ,Rana Dipti P., “No Title,”
A Nov. Fuzzy Based Classif. Data Min. using
Fuzzy Discret., no. World Congress on Computer
Science and Information, 2008.
[15]
X. Y. et Al., “Combination Data Mining Models
with New Medical Data to Predict Outcome of
Coronary Heart Disease,” Proc. Int. Conf. Converg.
Inf. Technol., pp. 868–872, 2007.
[16]
Jankowski, J., Floege, J., Fliser, D., Boehm, M., &
Marx, N. (2021). Cardiovascular disease in chronic
kidney disease: pathophysiological insights and
therapeutic options. Circulation, 143(11), 1157-
1172
[17]
Eckel, R. H., Bornfeldt, K. E., & Goldberg, I. J.
(2021). Cardiovascular disease in diabetes, beyond
glucose. Cell Metabolism, 33(8), 1519-1545.
[18]
S. G. Manoj B, Kumar G, Ramesh G, “Emerging
risk factors for cardiovascular diseases: Indian
context.,” Indian J Endocrinol Metab, vol. 17, pp.
806–814, 2013.
[19]
N. S. and D. Singh, “Performance Evaluation of K-
Means and Hierarchal Clustering in Terms of
Accuracy and Running Time,” Dep. Comput. Sci.
Eng. Barkatullah Univ. Inst. Technol., no. Ph.D
Dissertation, 2012
[20]
A. A. Shinde, S. N. Kale, R. M. Samant, A. S.
Naik, and S. A. Ghorpade, “Heart Disease
Prediction System using Multilayered Feed
Forward Neural Network and Back Propagation
Neural Network,” Int. J. Comput. Appl., vol. 166,
no. 7, pp. 975– 8887, 2
[21]
M. C. and P. M. Franck Le Duff,
CristianMunteanb, “Predicting Survival Causes
After Out of Hospital Cardiac Arrest using Data
Mining Method,” Stud. Health Technol. Inform.,
vol. Vol. 107, no. 2, p. No. 2.
[22]
Chaudhary, A., & Garg, P. (2014). Detecting and
diagnosing a disease by patient monitoring system.
International Journal of Mechanical Engineering
And Information Technology, 2(6), 493-499.
[23]
Garg, P.,
Malik,
K.,
Raheja,
N.,
&
Garg, P. (2011).
Enhanced FP-growth algorithm.
International
Journal of Computational Engineering
and
Management, 12, 54-56.
[24]
P. Garg, A. Dixit and P. Sethi, "Ml-fresh: novel
routing protocol in opportunistic networks using
machine learning," Computer Systems Science and
Engineering, vol. 40, no.2, pp. 703–717, 2022.
[25]
Yadav, P. S., Khan, S., Singh, Y. V., Garg, P., &
Singh, R. S. (2022). A Lightweight Deep Learning-
Based Approach for Jazz Music Generation in
MIDI Format. Computational Intelligence and
Neuroscience, 2022.
[26]
Soni, E., Nagpal, A., Garg, P., & Pinheiro, P. R.
(2022). Assessment of Compressed and
Decompressed ECG Databases for Telecardiology
Applying a Convolution Neural Network.
Electronics, 11(17), 2708.
[27]
Chauhan, S., Singh, M., & Garg, P. (2021). Rapid
Forecasting of Pandemic Outbreak Using Machine
Learning. Enabling Healthcare 4.0 for Pandemics:
A Roadmap Using AI, Machine Learning, IoT and
Cognitive Technologies, 59-73.
[28]
Gupta, S., & Garg, P. (2021). An insight review on
multimedia forensics technology. Cyber Crime and
Forensic Computing: Modern Principles, Practices,
and Algorithms, 11, 27.
[29]
Shrivastava, P., Agarwal, P., Sharma, K., & Garg,
P. (2021). Data leakage detection in Wi-Fi
networks. Cyber Crime and Forensic Computing:
Modern Principles, Practices, and Algorithms, 11,
215.
[30]
Pustokhina, I. V., Pustokhin, D. A., Lydia, E. L.,
Garg, P., Kadian, A., & Shankar, K. (2021).
Hyperparameter search based convolution neural
network with Bi-LSTM model for intrusion
detection system in multimedia big data
environment. Multimedia Tools and Applications,
1-18.
[31]
Khanna, A., Rani, P., Garg, P., Singh, P. K., &
Khamparia, A. (2021). An Enhanced Crow Search
Inspired Feature Selection Technique for Intrusion
Detection Based Wireless Network System.
Wireless Personal Communications, 1-18.
[32]
Meenakshi, P. G., & Shrivastava, P. (2021).
Machine learning for mobile malware analysis.
Cyber Crime and Forensic Computing: Modern
Principles, Practices, and Algorithms, 11, 151.
[33]
Garg, P., Pranav, S., & Prerna, A. (2021). Green
Internet of Things (G-IoT): A Solution for

International Journal of Intelligent Systems and Applications in Engineering
IJISAE, 2023, 11(2s), 165–173 | 173
Sustainable Technological Development. In Green
Internet of Things for Smart Cities (pp. 23-46).
CRC Press.
[34]
Nanwal, J., Garg, P., Sethi, P., & Dixit, A. (2021).
Green IoT and Big Data: Succeeding towards
Building Smart Cities. In Green Internet of Things
for Smart Cities (pp. 83-98). CRC Press.
[35]
Garg, P., Dixit, A., & Sethi, P. (2021, May). Link
Prediction Techniques for Opportunistic Networks
using Machine Learning. In Proceedings of the
International Conference on Innovative Computing
& Communication (ICICC).
[36]
Beniwal, S., Saini, U., Garg, P., & Joon, R. K.
(2021). Improving performance during camera
surveillance by integration of edge detection in IoT
system. International Journal of E-Health and
Medical Communications (IJEHMC), 12(5), 84-96.
[37]
Garg, P., Dixit, A., & Sethi, P. (2021, April).
Opportunistic networks: Protocols, applications &
simulation trends. In Proceedings of the
International Conference on Innovative Computing
& Communication (ICICC).
[38]
Garg, P., Dixit, A., & Sethi, P. (2021). Performance
comparison of fresh and spray & wait protocol
through one simulator. IT in Industry, 9(2).
[39]
Gupta, M., Garg, P., & Agarwal, P. (2021). Ant
Colony Optimization Technique in Soft
Computational Data Research for NP-Hard
Problems. In Artificial Intelligence for a
Sustainable Industry 4.0 (pp. 197-211). Springer,
Cham.
[40]
Malik, M., Singh, Y., Garg, P., & Gupta, S. (2020).
Deep Learning in Healthcare system. International
Journal of Grid and Distributed Computing, 13(2),
469-468.
[41]
Gupta, M., Garg, P., Gupta, S., & Joon, R. (2020).
A Novel Approach for Malicious Node Detection in
Cluster-Head Gateway Switching Routing in
Mobile Ad Hoc Networks. International Journal of
Future Generation Communication and
Networking, 13(4), 99-111.
[42]
Garg, P., Dixit, A., Sethi, P., & Pinheiro, P. R.
(2020). Impact of node density on the qos
parameters of routing protocols in opportunistic
networks for smart spaces. Mobile Information
Systems, 2020.
[43]
Gupta, A., Garg, P., & Sonal, Y. S. (2020). Edge
Detection Based 3D Biometric System for Security
of Web-Based Payment and Task Management
Application. International Journal of Grid and
Distributed Computing, 13(1), 2064-2076.
[44]
Dixit, A., Garg, P., Sethi, P., & Singh, Y. (2020,
April). TVCCCS: Television Viewer’s Channel
Cost Calculation System On Per Second Usage. In
IOP Conference Series: Materials Science and
Engineering (Vol. 804, No. 1, p. 012046). IOP
Publishing.
[45]
Sethi, P., Garg, P., Dixit, A., & Singh, Y. (2020,
April). Smart number cruncher–a voice based
calculator. In IOP Conference Series: Materials
Science and Engineering (Vol. 804, No. 1, p.
012041). IOP Publishing.
[46]
S. Rai, V. Choubey, Suryansh and P. Garg, "A
Systematic Review of Encryption and Keylogging
for Computer System Security," 2022 Fifth
International Conference on Computational
Intelligence and Communication Technologies
(CCICT), 2022, pp. 157-163, doi:
10.1109/CCiCT56684.2022.00039.
[47]
L. Saraswat, L. Mohanty, P. Garg and S. Lamba,
"Plant Disease Identification Using Plant Images,"
2022 Fifth International Conference on
Computational Intelligence and Communication
Technologies (CCICT), 2022, pp. 79-82, doi:
10.1109/CCiCT56684.2022.